26 THE CONDOR Vol. XXIII 
mens, where two or more subspecies may be concerned, on the basis of manifest 
characters irrespective of the probability of immediate blood relationships. In 
other words, an extreme of individual variation in one subspecies may be diag- 
nosed as belonging to.another subspecies. Great danger of coming to wrong con- 
clusions in regard to distribution in general and routes of migration in particu- 
lar will inevitably result from such interpretation. Far better to leave migrant 
examples, of equivocal appearance, undetermined as to subspecies. In the case 
of the Curlews in question the subspecifie distinctions claimed are at best only 
average, and the characters involved pertain to only one or two dimensions. 
How, then, can Oberholser, or anyone else, say of the majority of individuals 
taken in migration (when there are no geographic probabilities to regard) which 
are americanus and which are occidentalis? And yet Oberholser lists every spe- 
camen under one head or the other! 
As to Ridgway’s measurements, he gives his own, of 21 specimens of amer- 
tcanus and of a somewhat ‘‘larger’’ number of occidentalis (so stated [ footnote, 
p. 394], but not definitely specified because measurement sheets had ‘‘been mis- 
laid’’). Averages and extremes are presented. It seems quite likely that Ridg- 
way simply followed Oberholser’s lead as to the validity oi the separation. 
I would now eall attention to the accompanying graphic exhibition of the 
measurements of the Morro series in comparison with the measurements of the 
two alleged races as given by Oberholser and by Ridgway. The dimensions given 
are of wing, tail, tarsus and exposed culmen; in each case the dimensions are 
given graphically, natural size, that is, on a scale of 1 to 1. All the measure- 
ments of the Morro series were taken by myself. The figures for the graphs were 
checked back and forth with the kind assistance of Mr. Halsted G. White; and 
the graphs themselves were drawn and lettered by Mrs. St. E. Abernathy. An 
enquiry as to the meaning of these graphs (figs. 5 and 6) is in order, and some 
very curious things come of it. 
As to wing.—It is to be seen that, while Oberholser’s and Ridgway’s averages 
for americanus and occidentalis do not coincide, the amount of difference in- 
volved is close, namely, 6 and 5 percent, respectively, in males and 7 and 9 per- 
cent in females. Individual variation is such that overlapping of measurements 
takes place broadly in all cases except that of Oberholser’s males of americanus 
and occidentalis, between the nearest extremes of which there is a hiatus of about 
6mm. Ridgway’s longest winged occidentalis is within one millimeter of being 
as long as his longest americanus. | 
Referring to the Morro series, it is to be seen that in the males the average 
falls with occidentalis. Yet with the females it falls nearer americanus. (Some 
one may suggest that here is an instance of differential sex migration!) The 
individuals will be seen to be scattered along pretty evenly, without any obvious 
tendency to bi-modal grouping. 7 
As to tail.—The differences here between the measurements given of ameri- 
Canus and occidentalis vary from 6 to 14 percent. The amount of overlapping of 
extremes 18 very irregular. The Morro series, both males and females, will be 
seen to fall with occidentalis ; minima show even lesser figures. As Ridgway has 
pointed out (tom. cit., p. 391, footnote), there is a likelihood that differences 
due to method of taking the length of tail are involved. 
As to tarsus.—The percentages of difference between averages of americanus 
and occidentalis, as given by Oberholser and by Ridgway, and for the two sexes, 
range from 6 percent to as much as 17 percent. Yet there is fully 50 percent 
7 " si 
